Too good to be true

The CRFB's plan to save Social Security

The Committee for a Responsible Federal Budget (CRFB) has put forth an excellent plan to save Social Security, featuring a $100,000 benefit cap. The plan is so good that I see almost no prospect for it ever being enacted by our Congress, an institution that has fallen to a sadly dysfunctional state. In this post, I’ll discuss why I like the plan and then describe the type of far inferior plan likely to eventually be adopted.

The CRFB’s proposal is essentially a progressive consumption tax, although it won’t look like that to the average person. I cannot teach an entire course in public finance theory in a blog post, but the essence of a consumption tax is as follows:

Imagine a world where people can either spend $6000/month on consumption today, or $12,000/month on consumption in 20 years, by saving their incomes. Now assume you impose a 33.3% tax in that world, which takes away a third of the public’s resources for consumption. With a pure consumption tax, your choice is now $4000/consumption today or $8000 consumption in 20 years.

Notice that the “terms of trade” have not changed, in both cases, the opportunity cost of a dollar spent on consumption today is foregoing two dollar’s consumption in 20 years. A consumption tax is a tax that does not change the relative price of current and future consumption. In a sense, all taxes are consumption taxes, as the burden of any tax is its impact on a person’s lifetime consumption. However, economists use the term “consumption tax” to refer specifically to taxes that treat current and future consumption equally. An income tax punishes savers and hence is not a consumption tax.

I’m pretty sure that most people don’t understand this concept, as I often see commenters say things like “we should tax consumption, not labor.” Actually, a labor tax is a consumption tax. Indeed, these three taxes are all equivalent consumption taxes, in the long run:

1. A 20% VAT

2. A 20% payroll tax on wages

3. A 20% income tax with unlimited ability to put savings into a 401k plan, and no mandatory date of withdrawal from the 401k. (Funds borrowed for consumption are also taxed.)

Another misconception is that there is a big difference between benefit cuts and tax increases. Not so. There’s essentially no difference between cutting benefits of a wealthy Social Security recipient by $1000 and not cutting their benefits at all but instead taxing that person an extra $1000. The CRFB plan that I will discuss is generally framed as a “spending cut”, but it’s essentially a progressive consumption tax on high end Social Security recipients.

The CRFB plan is described as a $100,000 cap on the total amount of annual Social Security benefits that a household can receive, but the details are somewhat more complicated. That headline $100,000 cap applies to the official retirement age of 67, but as you probably know a retiree’s benefit level depends on when they retire. In order to prevent the cap from encouraging early retirement, they make it depend on age:

The $100,000 SFL would be adjusted based on marital status and collection age. A single person collecting at the NRA would face a $50,000 limit. A couple in which both spouses began collecting benefits at age 70 would face a $124,000 limit, reflecting the 24% delayed retirement credit. A couple with both spouses collecting at age 62 would face a $70,000 limit, reflecting the 30% early retirement actuarial reduction. Different claiming ages would result in a blended limit.

They present three ways in which the cap could be adjusted over time:

The SFL could be indexed over time in a variety of ways. For this analysis, Jason DeBacker of the Open Research Group modeled three options – a $100,000 limit indexed to inflation, a limit frozen in nominal terms at $100,000 for 20 years and then indexed to average wage growth, and a limit frozen at $100,000 for 30 years before being indexed to wage growth. The Inflation-Indexed SFL could also switch to wage indexing after a specified number of years.

Because wage inflation runs slightly over 1% higher than price inflation, these approaches have different long run implications for how the cap evolves over time:

I’ll probably be dead by 2046, so I’d be better off with the inflation adjusted cap (blue line.) But the basic idea is so good I’d be thrilled with any of the three versions.

These proposals are not enough to fully save Social Security. Thus they also propose a tax increase, and again the proposal is almost too good to be true:

This Trust Fund Solutions Initiative white paper suggests a new alternative – replacing the employer side of the payroll tax with a flat Employer Compensation Tax (ECT) on all employer compensation costs. While workers would continue to pay payroll taxes, employers would instead pay an ECT on all wages (with no tax cap) and all fringe benefits such as employer-sponsored insurance and stock options.

Not only is this a progressive consumption tax (relative to the current tax), it also reduces the distortion of current wage taxes that exempt health insurance benefits. That tax break has been an important factor driving up health care costs. (The employee side FICA should also tax health care benefits, as should the personal income tax.)

Part 2: The ant and the grasshopper

Unfortunately, the CRFB plan seems too good to be true, and I expect Congress to implement something far worse. In order to understand why, consider two neighbors that both spent their careers in upper-middle class jobs making close to the Social Security taxable maximum (currently $184,500). Both retire as single people entitled to roughly $50,000/year in benefits. Both would see their benefits capped in nominal terms, which means their real benefit levels would decline over time.

But these two neighbors differ in one very important way. Smith was a high spender who would buy the latest BMW, while Jones was a high saver who always bought used cars. Smith saved very little while Jones maxed out his 401k plan.

Now Smith starts whining to his congressman that the proposed cap is unfair. It should only apply to “the wealthy”. His neighbor Jones is now pulling $100,000/year out of his 401k and doesn’t “need” his Social Security benefit to rise with inflation. “Please make the cuts depend on income levels, not benefit levels.” Because America has far more grasshoppers than ants, Congress listens to the whiners and applies benefit cuts only to those with high current incomes, not those with high lifetime wage incomes. They punish savers and reward spendthrifts.

Why do I believe this will occur? Why am I so cynical? Because this is what Congress has been doing for the past 113 years. As a result of numerous policies, savers are effectively taxed at a higher rate than those who don’t save, which means that future consumption has become more expensive relative to current consumption. But it doesn’t look that way to the average person.

People focus on the fact that those who are currently wealthy have more resources than the less wealthy, even when the gap is 100% due to the less thrifty person choosing to spend at an earlier stage of their lives. In my thought experiment, the two neighbors were equally wealthy in the only way that matters—they had equal lifetime resources to allocate to consumption and simply choose to do so at different points in time. Smith consumed when he was young enough to enjoy it, and Jones foolishly waited until he was old, wrongly imagining that he could still get a thrill out of life at age 70.

At one time you might have expected the GOP to champion the interests of high savers, but those days are probably gone. The GOP coalition is trending toward low savers and the party is becoming increasingly “populist” on economic matters.

[Full disclosure: I’ve never bought a new car in my entire life and I always maxed out my 401k contributions. I’m the foolish miser.]

Part 3. Charity isn’t what you think it is

There’s an ongoing debate over how much money billionaires ought to donate to charity. Unfortunately, most people miss the point. The issue isn’t charity vs. investment; it is consumption vs. non-consumption. A charitable person is an individual that doesn’t consume much relative to his or her wealth. If you wish to consider heirs, you might say a charitable person is someone who ensures that he or she and all their future heirs consume only a modest portion of their current wealth.

But you can also argue that a charitable person is someone that maximizes their wealth, given the share of that wealth that they intend to use for consumption (both they and their heirs.) A wealthy person can become more charitable by reducing long run family consumption of wealth from 40% to 30%, but also by increasing their total stock of wealth and keeping the consumption share at 40%.

I have no idea what Elon Musk intends to do with his wealth, but you can make an argument that the most charitable use of his current resources is to build up even more wealth, and then eventually donate most of it to a good cause. But you can also make a good argument that Bill Gates is doing the right thing by donating a very high share of his wealth to various causes that he considers to be effective forms of altruism. Either approach is defensible, and the best path partly depends on subjective estimates of how much more wealth can be generated via productive investments.

As Matt Yglesias recently pointed out, those billionaires that consume a large share of their wealth are the actual problem. I don’t wish to sound like a scold, as I undoubtedly consume more than optimal from the perspective of a utilitarian like Peter Singer. If I were a billionaire I’d probably live in a very expensive mansion in coastal California. Like most people, I’m at least somewhat selfish. If I’m pointing fingers, then I’m including myself. But as a purely factual matter, more personal consumption comes at an opportunity cost, what else could have been done with those resources? Let’s just admit that most people are a mix of selfishness and altruism.

As an aside, in my view the hardest problem is figuring out how to be effectively altruistic, if you have decided that this is the path you’d like to take. As Tyler Cowen recently suggested, billionaires often end up funding causes of questionable value. I’m not trying to use this post to provide any sort of grand theory of altruism. Rather my point is much more basic:

To be altruistic is to forego consumption for you and your heirs. That’s it.

PS. This post has only examined one aspect of the Social Security problem and there is much more that could be said on the issue. Thus, it would have been better if the Social Security system had been fully funded from the beginning. And I’d prefer the wage tax (FICA) fall 100% on employees, as this would make for more informed voters. But I understand that these ideas are politically infeasible in the US, at least at the moment, and hence view the CRFB plan as a pragmatic compromise.

PPS. Steven Landsburg has a nice defense of Scrooge:

In this whole world, there is nobody more generous than the miser—the man who could deplete the world’s resources but chooses not to. The only difference between miserliness and philanthropy is that the philanthropist serves a favored few while the miser spreads his largess far and wide.

If you build a house and refuse to buy a house, the rest of the world is one house richer. If you earn a dollar and refuse to spend a dollar, the rest of the world is one dollar richer—because you produced a dollar’s worth of goods and didn’t consume them.

Who exactly gets those goods? That depends on how you save. Put a dollar in the bank and you’ll bid down the interest rate by just enough so someone somewhere can afford an extra dollar’s worth of vacation or home improvement. Put a dollar in your mattress and (by effectively reducing the money supply) you’ll drive down prices by just enough so someone somewhere can have an extra dollar’s worth of coffee with his dinner.

What did Bastiat say? “That which is seen, and that which is not seen.”